Improved RSSI Distribution for Indoor Localization Application based on Real Data Measurements
Rihab Souissi, Ilef Ktata, Salwa Sahnoun, Ahmad Fakhfakh, Faouzi Derbel
Abstract
The critical importance of accurate sensor node location data is highlighted by the rapid growth of wireless sensor network technology. This requirement has led to the development of location techniques. Among these techniques, Received Signal Strength Indication (RSSI) is the most widely used method due to its low cost. However, distance estimation using the path loss model of RSSI values is a challenging task. Accurate determination of critical coefficients in an indoor environment for an indoor positioning system, including various parameters of the wireless communication channel, significantly improves the overall positioning accuracy. The determination of the RSSI distribution and the path loss exponent η becomes critical in nonline-of-sight conditions. In this paper, we present an enhancement of RSSI distribution for indoor localization applications using real data measurements. This method aims to estimate the distances between fixed anchors and the mobile target for self-localization based on fitting the collected data with a combination with the Cauchy distribution. The proposed RSSI extraction algorithm presents an accuracy of about 80.73 % and 1.82 m as RMSE for distance estimation. The results of the indoor localization application using this algorithm to extract distances achieve an average error equal to 1.52 m. The simulations are implemented in the objective modular network testbed in the C++ (OMNeT++) discrete event simulator.